288 Proceedings of the Royal Physical Society. 
A. von Danckelman, writing of the Congo at Vivi, remarks 
that the temperature at the surface varied only some tenths 
of a degree (C.), whether in mid-current or at the banks. 
Dr W. Griffith, in the case of the Brahmaputra at Sadiya, 
found scarcely any difference between the temperatures of 
the water near the banks and some distance out in the 
stream. There is no doubt that in the instances of large 
swift rivers ike the Congo and the Brahmaputra there would 
be scarcely any marginal heating, and the same remark 
would apply to rapid rivers of much smaller dimensions. 
In the case of the Thames, usually a sluggish river, as I 
should observe, I found that the marginal shallow waters 
capable of being superheated to any marked extent lay out 
of the current, as in the eddies or backwaters, where the 
water is sometimes nearly still. They extended only a few 
feet from the shore, and had a depth only of a few inches, 
forming in fact a very insignificant proportion of the river’s 
bulk. In spring these marginal shallows were warmed two 
or three degrees (F.) above the temperature of the mass of 
the river when the day was sunny—during a hot summer 
afternoon the difference was from four to six degrees. Thus 
on a warm July day, when the temperature in mid-stream 
was 70° (F.), the thermometer, placed in water 13 inch deep 
and a foot from the shore, registered 76°; 2 feet farther out, 
in a depth of 34 inches, it showed 73°°5; and some 6 feet 
from the shore the temperature was nearly that of mid- 
stream. These marginal shallows had in summer a much 
sreater daily range of temperature and a higher daily mean 
than the bulk of the river possessed. With an excess of 
from four to six degrees during the day, they became one or 
two degrees cooler at night. Thus between the afternoon of 
July 7th and the following sunrise, whilst the air in the 
shade fell in temperature from 86° to 65° (F.), and the water 
in mid-stream from 71°8 to 71°11, the marginal shallows, 
where they were 2 inches deep, cooled from 75°:5 to 69°5, 
their mean of 72°°5 being one degree higher than that of the 
‘centre of the stream. These remarks apply only to the 
marginal shallows out of the current. Let these shallows be 
more or less in the current, and the difference between their 
